A number of the QA tests fail


  • Guest -
Hi,

I compiled nwchem on a Linux workstation which has 2 Intel(R) Xeon(R) CPU X5650 @ 2.67GH processors. I compiled the code for architecture LINUX. For example, the ptatom test fails and when I look at it the first significant difference occurs early on:

ok.out:

  Symmetry analysis of molecular orbitals - initial alpha
-------------------------------------------------------


!! scf_movecs_sym_adapt:   14 vectors were symmetry contaminated

 Symmetry fudging

!! scf_movecs_sym_adapt:    2 vectors were symmetry contaminated

 Numbering of irreducible representations:

    1 a1g         2 a1u         3 a2g         4 a2u         5 eg
6 eu 7 t1g 8 t1u 9 t2g 10 t2u

 Orbital symmetries:

    1 a1g         2 t1u         3 t1u         4 t1u         5 t2g
6 t2g 7 t2g 8 eg 9 eg 10 a1g
11 t1u 12 t1u 13 t1u 14 a1g 15 t2g
16 t2g 17 t2g 18 eg 19 eg 20 a1g

while out gives:

!! scf_movecs_sym_adapt:   15 vectors were symmetry contaminated

 Symmetry fudging

!! scf_movecs_sym_adapt:    3 vectors were symmetry contaminated

 Numbering of irreducible representations:

    1 a1g         2 a1u         3 a2g         4 a2u         5 eg
6 eu 7 t1g 8 t1u 9 t2g 10 t2u

 Orbital symmetries:

    1 a1g         2 t1u         3 t1u         4 t1u         5 eg
6 eg 7 t2g 8 t2g 9 t2g 10 a1g
11 t1u 12 t1u 13 t1u 14 a1g 15 eg
16 eg 17 t2g 18 t2g 19 t2g 20 a1g


and after that the differences continue in the energy etc. Comments? More information needed?

Thanks

Frans Trouw

Forum Vet
We'll need to see the whole output file and need some more information about the way you compiled the code. Is this a 32-bit or 64-bit platform?


Forum >> NWChem's corner >> Running NWChem